COMPARZ Post Hoc Analysis: Characterizing Pazopanib Responders With Advanced Renal Cell Carcinoma

COMPARZ Post Hoc Analysis: Characterizing

Pazopanib Responders With Advanced Renal Cell

Carcinoma

Cora N. Sternberg,

1

Robert J. Motzer,

2

Thomas E. Hutson,

3

Toni K. Choueiri,

4

Christian Kollmannsberger,

5

Georg A. Bjarnason,

6

Paul Nathan,

7

Camillo Porta,

8

Viktor Grünwald,

9

Luca Dezzani,

10

Jackie Han,

10

Nizar M. Tannir

11

Abstract

This post hoc analysis of the COMPARZ study (pazopanib, n[ 557; sunitinib, n [ 553) supported similar ef-ficacy of first-line pazopanib and first-line sunitinib treatment in advanced renal cell carcinoma. Patients who required dose modifications because of toxicity received higher cumulative doses with longer time of treat-ment and had significantly better objective response rate, progression-free survival, and overall survival than patients with minimal toxicity.

Background: The phase III COMPARZ study showed noninferior efficacy of pazopanib versus sunitinib in advanced renal cell carcinoma. In this COMPARZ post hoc analysis we characterized pazopanib responders, patient subgroups with better outcomes, and the effect of dose modification on efficacy and safety. Patients and Methods: Patients were randomized to pazopanib 800 mg/d (n¼ 557) or sunitinib 50 mg/d, 4 weeks on/2 weeks off (n ¼ 553). Secondary end points included time to complete response (CR)/partial response (PR); the proportion of patients with CR/PR10 months and progression-free survival (PFS)10 months; efficacy in patients with baseline metastasis; and logistic regression analyses of patient characteristics associated with CR/PR10 months. Median PFS, objective response rate (ORR), and safety were evaluated in patients with or without dose reductions or interruptions lasting7 days. Results: Median time to response was numerically shorter for patients treated with pazopanib versus sunitinib (11.9 vs. 17.4 weeks). Similar percentages of pazopanib and sunitinib patients had CR/PR10 months (14% and 13%, respectively), and PFS10 months (31% and 34%, respectively). For patients without versus with adverse event (AE)-related dose reductions, median PFS, median overall survival, and ORR were 7.3 versus 12.5 months, 21.7 versus 36.8 months, and 22% versus 42% (allP < .0001) for pazopanib, and 5.5 versus 13.8 months, 18.1 versus 38.0 months, and 16% versus 34% (all P < .0001) for sunitinib; results were similar for dose interruptions. Conclusion: Dose modifications when required because of AEs were associated with improved efficacy, suggesting that AEs might be used as a surrogate marker of adequate dosing for individual patients.

Clinical Genitourinary Cancer, Vol. 17, No. 6, 425-35 ª 2019 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords: First-line, Sunitinib, Tyrosine kinase inhibitor, VEGF, VEGFR Clinical Trial Identifier: NCT00720941.

Study data are available upon reasonable request to Jackie Han atjackie.han@novartis. com.

1_{Weill Cornell Medicine, Hematology/Oncology, New York, NY} 2_{Memorial Sloan Kettering Cancer Center, New York, NY} 3_{Texas Oncology/Baylor Sammons Cancer Center, Dallas, TX} 4

Department of Medical Oncology, Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA

5

Division of Medical Oncology, The University of British Columbia, BCCA Van-couver Cancer Centre, VanVan-couver, British Columbia, Canada

6_{Division of Medical Oncology, Sunnybrook Odette Cancer Centre, University of}

Toronto, Toronto, Ontario, Canada

7_{Department of Medical Oncology, Mount Vernon Cancer Center, Northwood,}

United Kingdom

8_{Medical Oncology, I.R.C.C.S. San Matteo University Hospital Foundation, Pavia,}

Italy

9_{University Hospital Essen, West-German Cancer Center, Internal Medicine (Tumor}

research) and Clinic for Urology, Essen, Germany

10

Novartis Pharmaceuticals Corporation, East Hanover, NJ

11

Department of Genitourinary Medical Oncology, The University of Texas M.D. Anderson Cancer Center, Houston, TX

Submitted: Nov 14, 2018; Revised: Jan 18, 2019; Accepted: Jan 25, 2019; Epub: Feb 6, 2019

Address for correspondence: Cora N. Sternberg, MD, FACP, Englander Institute for Precision Medicine, Weill Cornell Medicine, Hematology/Oncology, Belfer Research Building, 413 East 69th Street, Room 1412, New York, NY 10021

Introduction

Pazopanib was approved as first-line treatment for advanced renal cell carcinoma (aRCC) based on the phase III VEG105192 study in which pazopanib significantly prolonged progression-free survival (PFS) compared with placebo (median, 9.2 vs. 4.2 months; P< .0001), and this benefit was observed in treatment-naive and cytokine pretreated patients.1The randomized phase III COMPARZ study demonstrated noninferior efficacy of first-line pazopanib versus sunitinib.2 The primary end point of non-inferior PFS with pazopanib versus sunitinib was met (8.4 vs. 9.5 months; hazard ratio [HR], 1.05; 95% confidence interval [CI], 0.90-1.22), and the secondary end points of objective response rate (ORR) and overall survival (OS) supported the comparable efficacy of the 2 agents in favorable- and intermediate-risk patient populations. Differences in the safety profile revealed that Grade 3/4 adverse events (AEs) and symptomatic AEs were more frequent with sunitinib compared with pazopanib, and most (11/ 14) health-related quality of life measures significantly favored pazopanib over sunitinib, a finding that was confirmed in the PISCES patient preference study.3_{Our objectives in this post hoc} analysis of COMPARZ were to characterize pazopanib responders and evaluate whether patient subpopulations achieved better outcomes. Furthermore, because of previous observations showing a relationship between pazopanib or sunitinib exposure and effi-cacy and safety,4,5 and recent attempts to improve sunitinib’s safety profile with alternative sunitinib dosing regimens,6-9 _an additional objective was to evaluate the effect of dose modifica-tions on efficacy and safety outcomes in COMPARZ.

Patients and Methods

Study Design

The COMPARZ study was an international randomized, open-label, noninferiority phase III trial.2 _{Briefly, 1110 patients with} clear-cell aRCC were randomized 1:1 to receive pazopanib (800 mg once daily; n¼ 557) or sunitinib (50 mg once daily for 4 weeks, followed by 2 weeks without treatment; n¼ 553) in 6-week cycles. The primary end point was PFS as assessed by independent review. The study was powered to demonstrate noninferiority of pazopanib versus sunitinib. Secondary end points included OS, safety, and quality of life. COMPARZ was approved by the institutional review board or ethics committee at each participating center and was conducted in accordance with the provisions of the Declaration of Helsinki and Good Clinical Practice guidelines. Additional post hoc analyses of COMPARZ are reported herein.

Response

Imaging (computed tomography or magnetic resonance imaging) for disease assessment, response, and evaluation according to the Response Evaluation Criteria in Solid Tumors version 1.0 was performed in the intention to treat (ITT) population at baseline, every 6 weeks until week 24, and then every 12 weeks thereafter.2

Safety

Adverse events were recorded and graded according to the Na-tional Cancer Institute Common Terminology Criteria for Adverse Events version 3.0.10

Statistical Analysis

The time to response (complete response [CR]/partial response [PR]) for pazopanib and sunitinib was compared using descriptive statistics. The proportion of patients with a response (CR/PR) or PFS duration10 months in the ITT population was summarized; this is longer than median PFS with pazopanib or sunitinib in the COMPARZ study (8.4 months and 9.5 months, respectively).2 Median PFS, median OS, and ORR were evaluated for patients with no, any, 1, and 2 dose reductions or dose interruptions lasting7 days. For PFS and OS, unadjusted HRs and 2-sided log rank P values were estimated for patients with no versus any dose reductions or dose interruptions lasting 7 days, and for ORR, Fisher’s exact test was used to compare patients with no versus any dose reductions or dose interruptions lasting7 days.

The proportion of patients with AEs of special interest (diarrhea, fatigue, hypertension, palmar-plantar erythrodysesthesia [PPE], he-matologic AEs, and liver enzyme elevations) in the safety population (patients who received1 dose of study drug) were summarized for patients with no, any, 1, or 2 dose reductions or dose in-terruptions lasting7 days. The most common (5% incidence) AEs leading to dose modifications in either treatment group were evaluated.

Logistic regression analyses were performed in patients with CR/ PR duration10 months and duration 18 months, using select demographic and baseline characteristics (Karnofsky Performance Status, number of metastatic sites, number of involved organs, and Memorial Sloan Kettering Cancer Center risk category). No ad-justments were made for multiple comparisons. Median PFS, me-dian OS, and ORR were calculated in patients with and without baseline bone, lung, and kidney metastasis from the ITT population.

Results

Efficacy and Response

Of the 171 (30.7%) pazopanib and 137 (24.8%) sunitinib pa-tients who achieved CR/PR (ORR), the median time to response was numerically shorter for pazopanib (11.9 weeks; 95% CI, 11.3-12.1) compared with sunitinib (17.4 weeks; 95% CI, 12.7-18.0;

Table 1). A similar percentage of pazopanib and sunitinib patients had a CR/PR response10 months (14% and 13%, respectively) and 18 months (6% and 7%, respectively;Table 1). A similar percentage of pazopanib and sunitinib patients also achieved a PFS duration 10 months (31% and 34%, respectively) and 18 months (14% and 15%, respectively).

Dose Modifications and Efficacy

Dose modifications occurred in similar proportions of patients in the pazopanib and sunitinib groups (seeSupplemental Table 1in the online version). None, any, 1, and2 dose reductions occurred in 56%, 44%, 27%, and 18% of patients with pazopanib and 49%, 51%, 29%, and 21% of patients with sunitinib. None, any, 1, and2 dose interruptions occurred in 56%, 44%, 25%, and 19% of patients with pazopanib and 51%, 48%, 25%, and 24% of pa-tients with sunitinib. For the pazopanib and sunitinib arms, papa-tients who underwent dose modifications had a lower median average daily dose, with most dose reductions occurring within thefirst 3 to

426

-

Clinical Genitourinary Cancer December 2019

6 months of treatment (see Supplemental Figure 1 in the online version). However, median average daily dose increased for patients in the pazopanib group who underwent 1 dose reduction, which might be because of the small number of patients remaining at this later time point (n¼ 20 at 2 years; n ¼ 1 at 3 years).

For the pazopanib and sunitinib arms, patients who underwent dose modification had a higher median cumulative dose compared with patients who underwent no dose modification, which is likely explained by the longer time of study treatment for these patients (Table 2). Median PFS for patients with no versus any dose re-ductions was 7.3 months (95% CI, 5.3-8.3) versus 12.5 months (95% CI, 10.9-15.0; HR, 1.693; 95% CI, 1.365-2.099; P< .0001) for pazopanib and 5.5 months (95% CI, 4.3-8.1) versus

13.8 months (95% CI, 11.1-16.4; HR: 1.872; 95% CI, 1.484-2.361; P< .0001) for sunitinib (Table 3andFigure 1). Median OS for patients with no versus any dose reductions was 21.7 months (95% CI, 18.1-24.7) versus 36.8 months (95% CI, 33.1-not esti-mable [NE]; HR, 2.095; 95% CI, 1.634-2.685; P< .0001) for pazopanib and 18.1 months (95% CI, 14.1-23.4) versus 38.0 months (95% CI, 31.5-NE; HR, 2.138; 95% CI, 1.663-2.749; P< .0001) for sunitinib (Table 3andFigure 2). ORR for patients with no versus any dose reductions was 22% (95% CI, 17.1%-26.4%) versus 42% (95% CI, 36.1%-48.4%; difference, 20.5%; 95% CI, 12.8%-28.2%; P < .0001) for pazopanib and 16% (95% CI, 11.9%-20.7%) versus 34% (95% CI, 28.0%-39.1%; difference, 17.3%; 95% CI, 10.2%-24.4%; P < .0001) for

Table 1 Time to Response (CR/PR) and Proportion of Patients Who Achieved CR/PR or PFS‡10 Months and ‡18 Months (ITT Population)

Pazopanib (n[ 557) Sunitinib (n[ 553)

Patients Who Achieved CR/PR, na 171 137

Time to Response (CR/PR), Weeksb

First quartile (95% CI) 6.0 (6.0-6.3) 11.6 (8.3-12.0)

Median (95% CI) 11.9 (11.3-12.1) 17.4 (12.7-18.0)

Third quartile (95% CI) 17.6 (12.9-18.1) 30.1 (23.1-35.9)

CR/PR Duration

10 Months, n/n (%) 76/557 (14) 74/553 (13)

Progressed or died, n/n (%) 38/76 (50) 33/74 (45)

Censored, follow-up ended, n/n (%) 13/76 (17) 20/74 (27)

Censored, follow-up ongoing, n/n (%) 25/76 (33) 21/74 (28)

18 Months, n/n (%) 34/557 (6) 38/553 (7)

Progressed or died, n/n (%) 8/34 (24) 13/38 (34)

Censored, follow-up ended, n/n (%) 6/34 (18) 10/38 (26)

Censored, follow-up ongoing, n/n (%) 20/34 (59) 15/38 (39)

PFS Duration, n (%)

10 Months 175 (31) 186 (34)

18 Months 79 (14) 85 (15)

Abbreviations: CI = con_{fidence interval; CR = complete response; ITT = intention to treat; PFS = progression-free survival; PR = partial response.}

a_{The time to response analysis was restricted to the subgroup of patients who experienced a CR/PR.} b_{Time to response is de}

fined as the time from the start of treatment until the first documented evidence of confirmed CR or PR, whichever comes first.

Table 2 Median Time on Treatment and Cumulative Dose (Safety Population)

Median Cumulative Dose,31000 mg (Range) Median Time of Study Treatment, Months (Range) Pazopanib (n[ 554) Sunitinib (n[ 548) Pazopanib (n[ 554) Sunitinib (n[ 548)

Dose Reduction(s)

None 134.4 (1.6-940.0) 4.2 (0.05-39.15) 5.6 (0-39) 3.7 (0-38)

Any 165.2 (9.6-804.0) 8.825 (0.80-32.25) 10.1 (1-40) 10.8 (1-38)

1 169.4 (9.6-804.0) 8.013 (0.80-29.05) 8.9 (1-40) 9.2 (1-38)

2 150.2 (21.2-750.2) 10.406 (1.713-32.25) 11.4 (1-39) 13.7 (2-37)

Dose Interruption(s)‡7 Days

None 116.8 (1.6-887.2) 2.85 (0.50-32.20) 5.1 (0-40) 2.3 (0-37)

Any 116.0 (15.2-940.0) 9.294 (0.80-39.15) 9.6 (1-39) 10.6 (1-38)

Table 3 PFS, OS, and ORR in the ITT Population by Dose Modification Group

Median PFS, Months (95% CI) Median OS, Months (95% CI) ORR, n (%) [95% CI]a

Pazopanib (n[ 554) Sunitinib (n[ 548) Pazopanib (n[ 554) Sunitinib (n[ 548) Pazopanib (n[ 554) Sunitinib (n[ 548) Dose Reduction(s)

None 7.3 (5.3-8.3) 5.5 (4.3-8.1) 21.7 (18.1-24.7) 18.1 (14.1-23.4) 67 (22) [17.1-26.4] 44 (16) [11.9-20.7]

Any 12.5 (10.9-15.0) 13.8 (11.1-16.4) 36.8 (33.1-NE) 38.0 (31.5-NE) 104 (42) [36.1-48.4] 93 (34) [28.0-39.1]

None versus any HR, 1.693 (95% CI,

1.365-2.099);P < .0001 HR, 1.872 (95% CI, 1.484-2.361);P < .0001 HR, 2.095 (95% CI, 1.634-2.685);P < .0001 HR, 2.138 (95% CI, 1.663-2.749);P < .0001 Difference: 20.5% (95% CI, 12.8%-28.2%);P < .0001 Difference: 17.3% (95% CI, 10.2%-24.4%);P < .0001 1 11.1 (8.3-13.5) 11.1 (10.2-13.8) 33.1 (27.2-NE) 30.3 (24.7-NE) 49 (33) [25.7-41.0] 50 (31) [23.9-38.2] 2 16.4 (11.1-18.6) 16.5 (11.5-19.3) NR (NE-NE) NR (34.9-NE) 55 (56) [45.8-65.3] 43 (37) [28.3-45.9]

Dose Interruption(s)‡7 Days

None 8.2 (5.5-8.3) 5.6 (5.4-8.2) 21.7 (17.8-26.0) 18.1 (14.2-23.2) 72 (23) [18.5-27.8] 46 (16) [12.1-20.8]

Any 12.6 (9.9-16.4) 13.8 (11.1-16.6) NR (31.6-NE) NR (32.1-NE) 99 (41) [34.6-46.9] 91 (34) [28.4-39.8]

None versus any HR: 1.648 (95% CI,

1.329-2.043);P < .0001 HR: 1.923 (95% CI, 1.529-2.418);P < .0001 HR: 1.959 (95% CI, 1.528-2.511);P < .0001 HR: 2.264 (95% CI, 1.762-2.909);P < .0001 Difference: 17.6% (95% CI, 9.8%-25.3%);P < .0001 Difference: 17.7% (95% CI, 10.5%-24.8%);P < .0001 1 8.3 (6.0-11.0) 11.0 (8.2-14.0) 31.6 (26.5-NE) 30.5 (23.7-NE) 42 (30) [22.8-38.1] 40 (30) [21.9-37.3] 2 16.7 (13.7-19.4) 16.6 (13.6-19.6) NR (36.8-NE) NR (34.9-NE) 57 (54) [44.8-63.8] 51 (39) [30.3-46.9]

Abbreviations: CI = confidence interval; HR = hazard ratio; ITT = intention to treat; NE = not estimable; NR = not reached; OS = overall survival; ORR = objective response rate; PFS = progression-free survival.

a_{Percentage was calculated using the number of patients in the corresponding dose modification group (seeSupplemental Table 1in the online version) as the denominator.}

COMPARZ

Post

Hoc

Analysis

of

Pazopanib

Responders

428

-Clinica lGenitourinar y Cancer December 2019

sunitinib (Table 3). Similarfindings were observed for patients who underwent dose interruptions of7 days’ duration (Table 3and

Figures 1and2), suggesting that patients requiring dose

modifica-tions because of AEs were more likely to respond and to have a longer PFS and OS.

Predictors of Efficacy and Response

Logistic regression analyses did not identify baseline patient characteristics significantly associated with response in either the pazopanib or sunitinib groups when comparing patients with a CR/ PR duration of10 versus <10 months (seeSupplemental Table 2

in the online version). Median PFS and median OS in patients with

pazopanib and sunitinib (seeSupplemental Table 3in the online version). ORR was significantly higher for pazopanib versus suni-tinib in patients with baseline lung metastasis (36% vs. 28%; P¼ .008).

Safety and Dose Modifications

Select AEs (diarrhea, fatigue, hypertension, PPE, hematologic AEs, and alanine aminotransferase [ALT]/aspartate aminotransferase [AST] elevations) were more frequent in patients who underwent dose reductions or interruptions (seeSupplemental Figure 2in the online version). Consistent with the primary COMPARZ analysis,2 PPE and hematologic AEs occurred more frequently with sunitinib

Figure 1 KaplaneMeier Plots of Progression-Free Survival in Patients With and Without Dose Reductions With Pazopanib (A) and Sunitinib (B), and in Patients With and Without Dose Interruptions With Pazopanib (C) and Sunitinib (D). Error Bars Indicate 95% CIs

0.0

Time Since Randomization (Months)

Pr ogr e ssion-fr ee Survival Pr obability 246 195 138 88 67 31 DR 22 10 6 1 0 0 4 8 12 16 20 24 28 32 36 40 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

1.0 _{No dose reduction (NoDR) (N = 308)} Any dose reduction (DR) (N = 246)

NoDR

Number of patients at risk for pazopanib:

308 166 107 48 38 30 24 9 7 0 0

0.0

Time Since Randomization (Months)

Pr ogr e ssion-fr ee Survival Pr obability 277 233 175 108 82 51 NoDR 37 16 10 3 0 0 4 8 12 16 20 24 28 32 36 40 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

1.0 _{No dose reduction (NoDR) (N = 270)} Any dose reduction (DR) (N = 277)

DR

Number of patients at risk for sunitinib:

270 118 74 39 29 18 11 2 0 0 0

A

pazopanib and sunitinib, the incidence of AEs was higher with than without dose modification. The most common (10%) AEs lead-ing to dose modification with pazopanib were hypertension (13%), fatigue (12%), and diarrhea (11%), and the most common AEs leading to dose modification with sunitinib were fatigue (15%), PPE (12%), thrombocytopenia (12%), and diarrhea (10%) (see

Supplemental Table 4in the online version).

Discussion

This post hoc analysis of COMPARZ demonstrated that time to response was excellent with both drugs, although numerically shorter with pazopanib compared with sunitinib; the proportion of patients with a (10 months) response was similar for pazopanib (14%) versus sunitinib (13%); and patients who experienced clinical

benefit from pazopanib or sunitinib were more likely to have experienced AEs requiring dose modifications.

Time to response for pazopanib in this post hoc analysis is consistent withfindings from the trial that led to approval of first-line pazopanib for aRCC, in which the median time to response was also 11.9 weeks according to independent review.1 With sunitinib, the median time to response in this post hoc analysis was 17.4 weeks, and although this cannot be compared directly with the sunitinib pivotal trial, a pooled analysis of 1059 meta-static renal cell carcinoma (RCC) patients treated with sunitinib across 6 clinical trials (including the pivotal trial) found a median time to response of 10.6 weeks.11_{The patient population in the} pooled clinical trial analysis was treatment-naive or cytokine pretreated, treated with sunitinib 50 mg/d (4/2 schedule) or

Figure 1 Continued

0.0

Time Since Randomization (Months)

Pr ogr e ssion-fr ee Survival Pr obability 243 188 133 84 64 34 NoDI 26 12 7 1 0 0 4 8 12 16 20 24 28 32 36 40 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

1.0 _{No dose interruption (NoDI) (N = 311)} Any dose interruption (DI) (N = 243)

DI

Number of patients at risk for pazopanib:

311 173 112 52 41 27 20 7 6 0 0

0.0

Time Since Randomization (Months)

Pr ogr e ssion-fr ee Survival Pr obability 267 221 166 106 84 54 NoDI 37 14 8 3 0 0 4 8 12 16 20 24 28 32 36 40 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

1.0 _{No dose interruption (NoDI) (N = 280)} Any dose interruption (DI) (N = 267)

DI

Number of patients at risk for sunitinib:

280 130 83 41 27 15 11 4 2 0 0

C

D

COMPARZ Post Hoc Analysis of Pazopanib Responders

continuous sunitinib 37.5 mg/d,11 _{and were thus a more} het-erogeneous patient population than in COMPARZ.

In this post hoc analysis of the COMPARZ trial, pazopanib and sunitinib were associated with a similar proportion of patients who had a response duration 10 months, as well as associated with comparable median PFS and median OS in patients with baseline bone, lung, and kidney metastasis. The noninferior efficacy and differentiated safety profile of first-line pazopanib and sunitinib treatment is supported by large real-world analyses.12-14

Patients who underwent dose modifications because of AEs continued therapy for longer periods of time, had significantly improved PFS, OS, and ORR, received a higher median cumulative

modification group, select AEs (PPE and hematologic AEs) were more common with sunitinib compared with pazopanib and liver enzyme elevations were more common with pazopanib, consistent with the primary analysis.2 This highlights the need for better therapy management for these patients, which might include dose reduction and treatment interruptions, which could ultimately lead to improved clinical outcomes.

Although other analyses of clinical studies support that increased exposure to pazopanib and sunitinib is associated with improved clinical outcomes, the current post hoc analysis of COMPARZ extends this by suggesting that dose modifications when required because of toxicity do not compromise efficacy. Further, this

anal-Figure 2 KaplaneMeier Plots of Overall Survival in Patients With and Without Dose Reductions With Pazopanib (A) and Sunitinib (B), and in Patients With and Without Dose Interruptions With Pazopanib (C) and Sunitinib (D). Error Bars Indicate 95% CIs

0.0

Time Since Randomization (Months)

Overall Survival Pr obability 0 4 8 12 16 20 24 28 32 36 40 44 48 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

1.0 _{No dose reduction (NoDR) (N = 308)} Any dose reduction (DR) (N = 246)

246 244 213 189 166 142 NoDR

71

120 37 14 2 0 0

DR

Number of patients at risk for pazopanib:

308 277 245 195 161 132 103 71 45 14 1 0 0

0.0

Time Since Randomization (Months)

Overall Survival Pr obability 8 4 6 3 2 3 8 2 4 2 0 2 6 1 2 1 8 4 0 44 277 269 246 212 192 167 NoDR 0 0 2 5 4 1 9 0 4 1 2 0 40 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

1.0 _{No dose reduction (NoDR) (N = 270)} Any dose reduction (DR) (N = 277)

DR

Number of patients at risk for sunitinib:

270 232 185 142 121 102 85 57 24 8 1 0 0

A

clinical benefit from pazopanib and sunitinib compared with patients who experience minimal toxicity. Thus, dose reductions and dose interruptions are safe for patients who experience toxicity from pazopanib or sunitinib. A retrospective analysis of 2 pro-spective sunitinib trials similarly demonstrated improved median PFS and ORR in patients who underwent dose reduction because of AEs compared with patients who remained on the standard 50 mg/d, 4/2 schedule.15A real-world study of 591 metastatic RCC patients treated withfirst-line pazopanib or sunitinib in Italy also suggested that dose modifications when necessary for AEs do not compromise efficacy.16_{In contrast, a chart review of 10 oncology} centers in Europe (n¼ 291) found significantly shorter survival for aRCC patients who received a low relative dose intensity (RDI) of

pazopanib or sunitinib (RDI<0.7),17as might be expected based on pharmacokinetic data demonstrating a positive relationship be-tween pazopanib/sunitinib exposure and survival.4,5_{However, 19%} of patients in the European chart review had initiated treatment at a lower than standard dose, and thus lower doses received in this patient population were not all due to toxicity.17Finally, a recent study in the adjuvant setting showed that higher pazopanib levels were associated with improved disease-free survival and did not increase treatment discontinuations or Grade 3/4 AEs, with the exception of hypertension.18This highlights the important role of drug exposure on clinical outcomes; dose modifications should only be considered for patients who require this intervention because of AEs.

Figure 2 Continued

0.0

Time Since Randomization (Months)

Overall Survival Pr obability 0 4 8 12 16 20 24 28 32 36 40 44 48 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

1.0 _{No dose interruption (NoDI) (N = 311)} Any dose interruption (DI) (N = 243)

243 239 211 188 165 147 NoDI

74

120 38 15 1 0 0

DI

Number of patients at risk for pazopanib:

311 282 247 196 162 127 103 68 44 13 2 0 0

0.0

Time Since Randomization (Months)

Overall Survival Pr obability 8 4 6 3 2 3 8 2 4 2 0 2 6 1 2 1 8 4 0 40 44 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

1.0 _{No dose interruption (NoDI) (N = 280)} Any dose interruption (DI) (N = 267)

267 258 239 206 189 167 NoDI 0 9 1 9 4 6 9 0 4 1 1 0 DI

Number of patients at risk for sunitinib:

0 9 0 2 2 5 5 8 2 0 1 4 2 1 8 4 1 2 9 1 3 4 2 0 8 2 2 0

C

D

COMPARZ Post Hoc Analysis of Pazopanib Responders

The relationship between exposure and efficacy end points has been demonstrated for several approved vascular endothelial growth factor receptor (VEGFR) tyrosine kinase inhibitors (TKIs), including pazopanib and sunitinib.4,5,19Higher sunitinib exposure has been shown to be associated with longer time to progression and greater OS in a pooled clinical trial analysis of metastatic RCC patients.4Similarly, survival benefits have been shown for pazonib. In a large phase II trial of pazopanib in /metastatic RCC pa-tients, a steady-state trough concentration of 20.5

m

g/mL was identified as the cutoff associated with improved PFS and tumor shrinkage, and a relationship between increased pazopanib exposure and the frequency of AEs, such as hypertension and liver enzyme elevations, was shown.5Indeed, high interpatient variability in drug exposure is observed with VEGFR TKIs, such as pazopanib and sunitinib,20,21and patients achieving higher drug exposure are also more likely to experience toxicity as well as survival benefit. This higher toxicity might lead to dose reductions or interruptions, which might explain the superior clinical outcomes (PFS, OS, and ORR) in patients with dose modifications compared with patients without dose modifications.

With sunitinib, the safety profile at the approved dose of 50 mg/d for 4 weeks followed by 2 weeks off treatment (4/2 schedule) has led to investigation of alternative off-label dose schedules, such as the 2/ 1 schedule6,7and continuous 37.5 mg/d dosing (Renal EFFECT trial).8_{Although the Renal EFFECT study did not lead to a change} in practice, the 2:1 schedule for patients experiencing toxicity with the 4:2 schedule has been widely used. Outcomes from a single-arm phase II trial of individualized sunitinib dosing also support dose/ schedule individualization in patients experiencing AEs.9 In this individualized-dosing study, dose reductions and schedule changes were implemented in patients experiencing grade2 toxicity, and patients experiencing minimal toxicity received dose escalation (18.4%). The median PFS was 12.5 months and median OS was 38.5 months. The ORR (46.1%) and stable disease rate (38.5%) translated into a clinical benefit for 84.6% of patients with no decline in quality of life scores during therapy. Although multiple studies have assessed the effect of nonstandard intermittent dosing schedules with sunitinib to maintain therapeutic drug concentra-tions, prolong duration of therapy, minimize AEs, and/or maintain efficacy, we are unaware of any clinical studies prospectively investigating intermittent pazopanib dosing schedules in patients with aRCC. However, a recent preclinical study reported that a high-dose intermittent pazopanib dosing schedule was able to extend median OS in an animal model of advanced metastatic RCC resistant to continuous pazopanib,22 suggesting potential clinical utility for intermittent pazopanib dose scheduling for selected pa-tients with aRCC.

The association between tolerability and clinical outcomes underscores that clinical outcomes are not adversely affected in pa-tients with treatment-related AEs who undergo dose reductions and remain on therapy. This supports individual dosing titrated according to toxicity. Although patients without toxicity have worse outcomes, the consequences of this on dosing and treatment strategy are less clear. Dose reductions should only be applied following presentation

of treatment-related AEs dose reductions are not an intervention that improves long-term outcomes, but rather a necessity to keep patients on treatment. Whether patients who experience minimal toxicity should be dose escalated is a valid question that should be addressed by future studies. Initial observations suggest TKI dose escalation during treatment may be appropriate for selected patients with metastatic RCC. In a retrospective analysis of 25 patients whose disease pro-gressed during sunitinib treatment, 36% had a PR and 28% had stable disease for a median of 7.5 months after dose escalation.23In the phase II study of individualized sunitinib treatment previously discussed, 18.4% of patients were dose escalated.9_{Axitinib titration was} asso-ciated with improved response rates in a randomized phase II trial.24 In a retrospective analysis of 22 patients who received an escalated TKI dose (axitinib [17], sunitinib [3], pazopanib [2]) after progressive disease, 4 (22%) patients experienced a PR and 78% had a decreased disease burden after dose escalation.25Individualizing axitinib dose and treatment duration based on toxicity with planned breaks of therapy has been reported to be feasible and active.26

Limitations of this study are the post hoc, retrospective nature of the analyses. Furthermore, no adjustments were made for multiple comparisons, limiting the conclusions that can be drawn from the efficacy by baseline metastatic site data.

In summary, these results suggest that clinicians treating aRCC patients with sunitinib or pazopanib should reduce the dosage and/ or give treatment breaks if required because of AEs, which might allow patients to remain longer on treatment and continue to obtain clinical benefit. Differences revealed between first-line pazopanib and sunitinib may also aid treatment choice for clinicians, such as the shorter time to response and lower frequency of PPE and he-matologic AEs with pazopanib, and lower frequency of ALT/AST elevations with sunitinib.

Conclusion

In this post hoc analysis of the COMPARZ study, patients who required dose reductions and dose interruptions due to AEs expe-rienced longer time on treatment, received greater cumulative doses, and had significantly improved PFS, OS, and ORR compared with patients who did not require dose modifications. This indicates that dose modifications can be safely implemented without compro-mising pazopanib or sunitinib efficacy, and that AEs might be used as a surrogate marker of adequate dosing for individual patients.

Clinical Practice Points

 In the phase III COMPARZ study, first-line pazopanib was noninferior to first-line sunitinib with regard to efficacy in metastatic RCC; safety and quality of life profiles favored pazopanib.

 In this post hoc analysis of COMPARZ, a similar percentage of patients with pazopanib and sunitinib had a response duration (CR/PR or PFS)10 months.

 The median time to response was 11.9 weeks with pazopanib versus 17.4 weeks with sunitinib.

 Within both arms, patients with AE-related dose modifications had higher cumulative doses; longer time on treatment,

significantly improved PFS, OS, and ORR; and more frequent AEs versus patients with no dose modification.

 Thesefindings suggest that clinicians can safely alter pazopanib or sunitinib dosing because of AEs without compromising effi-cacy and that AEs might be used as a surrogate marker of adequate dosing for each patient.

Acknowledgments

Editorial assistance was provided by Chris Ontiveros, PhD (ApotheCom, New York, NY), and Julia Burke, PhD (ApotheCom, Auckland, New Zealand), and was funded by Novartis Pharma-ceuticals Corporation. The COMPARZ study was supported by GlaxoSmithKline Pharmaceuticals, and this post hoc analysis was sponsored by Novartis; pazopanib is an asset of Novartis AG as of March 2, 2015. The sponsor providedfinancial support for study conduction and preparation of the final report Additional support was provided by Memorial Sloan-Kettering Cancer Center Support Grant/Core Grant (P30 CA008748).

Disclosure

Cora N. Sternberg has received consulting and/or advisor fees from Bayer, BMS, Eisai, IPSEN, Merck, Novartis, Pfizer, and Roche-Genentech as well as research funding from Eisai, Exelixis, Pfizer, and Roche-Genentech. Robert J. Motzer has received consulting and/or advisor fees from Eisai, Exelixis, Genentech/ Roche, Merck, Novartis, and Pfizer, as well as research funding from BMS, Eisai, Exelixis Genentech/Roche, Novartis, and Pfizer. Thomas E. Hutson has received honoraria from Astellas, Bayer, GSK, Novartis, and Pfizer, consulting and/or advisor fees from Bayer, GSK, Novartis, and Pfizer; speakers bureau fees from Astellas, Bayer, GSK, Janssen, Novartis, and Pfizer; and research funding from Astellas, GSK, Janssen, Novartis, and Pfizer. Toni K. Choueiri has received consulting and/or advisor fees from Bayer, BMS, Eisai, GSK, Merck, Novartis, Pfizer, and Prometheus Labs, Inc, and research funding from AstraZeneca, BMS, Exelixis, GSK, Merck, Novartis, Pfizer, Roche, and Tracon. Christian Koll-mannsberger has received honoraria from BMS, Eisai, Ipsen, Novartis, and Pfizer, travel funding from Novartis, Pfizer and Sanofi, and consulting and/or advisor fees from Astellas Pharma, Bayer, BMS, Eisai, Ipsen, Novartis, Pfizer, and Seattle Genetics. Georg A. Bjarnason has received honoraria from BMS, Eisai, Ipsen, Novartis, and Pfizer, research funding from Novartis and Pfizer, travel funding from Novartis and Pfizer, and consulting and/or advisor fees from BMS, Eisai, Ipsen, Novartis, and Pfizer. Paul Nathan has received consulting, advisory board, and/or speakers bureau fees from AZ, BMS, Ipsen, Merck, MSD, Novartis, Pfizer, and Roche. Camillo Porta has received consulting and/or advisor fees from BMS, Eisai, EUSA Pharma, Ipsen, Janssen, Novartis, Peloton, and Pfizer, speakers bureau fees from BMS, Eisai, Ipsen, Novartis, and Pfizer, and research funding from Pfizer. Viktor Grünwald has received grants from AstraZeneca, BMS, MSD, and Pfizer, personal fees from AstraZeneca, BMS, Eisai, EUSA Pharma, Ipsen, MSD, Novartis, Pfizer, and Roche, and nonfinancial support from BMS, Ipsen, MSD, Novartis, and Roche. Luca Dezzani is employed by Novartis. Jackie Han is employed by and has stock or

ownership in Novartis. Nizar M. Tannir has received consulting and/or advisor fees from Argos, BMS, Calithera, Exelixis, Nektar, Novartis, and Pfizer, research funding from BMS, Epizyme, Exe-lixis, Miranti, and Novartis, and fees for travel, accommodation, and/or expenses from Argos, BMS, Calithera, Exilixis, Nektar, Novartis, and Pfizer.

Supplemental Data

Supplemental tables andfigures accompanying this article can be found in the online version athttps://doi.org/10.1016/j.clgc.2019. 01.015.

References

1.Sternberg CN, Davis ID, Mardiak J, et al. Pazopanib in locally advanced or metastatic renal cell carcinoma: results of a randomized phase III trial. J Clin Oncol 2010; 28:1061-8.

2.Motzer RJ, Hutson TE, Cella D, et al. Pazopanib versus sunitinib in metastatic renal-cell carcinoma. N Engl J Med 2013; 369:722-31.

3.Escudier B, Porta C, Bono P, et al. Randomized, controlled, double-blind, cross-over trial assessing treatment preference for pazopanib versus sunitinib in patients with metastatic renal cell carcinoma: PISCES study. J Clin Oncol 2014; 32:1412-8. 4.Houk BE, Bello CL, Poland B, Rosen LS, Demetri GD, Motzer RJ. Relationship between exposure to sunitinib and efficacy and tolerability endpoints in patients with cancer: results of a pharmacokinetic/pharmacodynamic meta-analysis. Cancer Chemother Pharmacol 2010; 66:357-71.

5.Suttle AB, Ball HA, Molimard M, et al. Relationships between pazopanib exposure and clinical safety and efficacy in patients with advanced renal cell carcinoma. Br J Cancer 2014; 111:1909-16.

6.Lee JL, Kim MK, Park I, et al. RandomizEd phase II trial of Sunitinib four weeks on and two weeks off versus Two weeks on and One week off in metastatic clear-cell type REnal clear-cell carcinoma: RESTORE trial. Ann Oncol 2015; 26:2300-5. 7.Bracarda S, Iacovelli R, Boni L, et al. Sunitinib administered on 2/1 schedule in

patients with metastatic renal cell carcinoma: the RAINBOW analysis. Ann Oncol 2015; 26:2107-13.

8.Motzer RJ, Hutson TE, Olsen MR, et al. Randomized phase II trial of sunitinib on an intermittent versus continuous dosing schedule asfirst-line therapy for advanced renal cell carcinoma. J Clin Oncol 2012; 30:1371-7.

9.Bjarnason GA, Knox JJ, Kollmannsberger CK, et al. The efficacy and safety of sunitinib given on an individualised schedule asfirst-line therapy for metastatic renal cell carcinoma: a phase 2 clinical trial. Eur J Cancer 2019; 108:69-77. 10. National Institutes of Health. National Cancer Institute. Common terminology

criteria for adverse events (CTCAE). Available at: https://ctep.cancer.gov/ protocolDevelopment/electronic_applications/ctc.htm. Updated March 1, 2018. Accessed August 8, 2019.

11.Molina AM, Lin X, Korytowsky B, et al. Sunitinib objective response in metastatic renal cell carcinoma: analysis of 1059 patients treated on clinical trials. Eur J Cancer 2014; 50:351-8.

12.Ruiz-Morales JM, Swierkowski M, Wells JC, et al. First-line sunitinib versus pazopanib in metastatic renal cell carcinoma: results from the International Met-astatic Renal Cell Carcinoma Database Consortium. Eur J Cancer 2016; 65:102-8. 13.Hirsch BR, Jiao X, Wilson T, et al. Comparative effectiveness of pazopanib and sunitinib asfirst-line therapy for patients with advanced/metastatic renal cell car-cinoma in a US community oncology setting (abstract 567). J Clin Oncol 2016; 34(suppl 2):567.

14.Lalani AA, Li H, Heng DY, et al. First-line sunitinib or pazopanib in metastatic renal cell carcinoma: the Canadian experience.Can Urol Assoc J 2017; 11:112-7. 15.Khosravan R, Huang X, Wiltshire R, Lechuga M, Motzer RJ. A retrospective analysis of data from two trials of sunitinib in patients with advanced renal cell carcinoma (RCC): pitfalls of efficacy subgroup analyses based on dose-reduction status (abstract 363). J Clin Oncol 2012; 30(suppl 5):363.

16.Iacovelli R, Cossu Rocca M, Galli L, et al. Clinical outcome of patients who reduced sunitinib or pazopanib duringfirst-line treatment for advanced kidney cancer. Urol Oncol 2017; 35, 541.e7-541.e13.

17.Porta C, Levy A, Hawkins R, et al. Impact of adverse events, treatment modi fi-cations, and dose intensity on survival among patients with advanced renal cell carcinoma treated withfirst-line sunitinib: a medical chart review across ten centers infive European countries. Cancer Med 2014; 3:1517-26.

18.Sternberg CN, Donskov F, Haas NB, et al. Pazopanib exposure relationship with clinical efficacy and safety in the adjuvant treatment of advanced renal cell carci-noma. Clin Cancer Res 2018; 24:3005-13.

19.Rini BI, Melichar B, Fishman MN, et al. Axitinib dose titration: analyses of exposure, blood pressure and clinical response from a randomized phase II study in metastatic renal cell carcinoma. Ann Oncol 2015; 26:1372-7.

20.Heath EI, Chiorean EG, Sweeney CJ, et al. A phase I study of the pharmacokinetic and safety profiles of oral pazopanib with a high-fat or low-fat meal in patients with advanced solid tumors. Clin Pharmacol Ther 2010; 88:818-23.

COMPARZ Post Hoc Analysis of Pazopanib Responders

21.Houk BE, Bello CL, Kang D, Amantea M. A population pharmacokinetic meta-analysis of sunitinib malate (SU11248) and its primary metabolite (SU12662) in healthy volunteers and oncology patients. Clin Cancer Res 2009; 15:2497-506.

22.Reguera-Nuñez E, Man S, Xu P, Kerbel KS. Preclinical impact of high dose intermittent antiangiogenic tyrosine kinase inhibitor pazopanib in intrinsically resistant tumor models. Angiogenesis 2018; 21:793-804.

23.Raphael J, Thawer A, Bjarnason GA. Sunitinib dose-escalation after dis-ease progression in metastatic renal cell carcinoma. Urol Oncol 2018; 36: 12.e1-6.

24.Rini BI, Melichar B, Ueda T, et al. Axitinib with or without dose titration for first-line metastatic renal-cell carcinoma: a randomised double-blind phase 2 trial. Lancet Oncol 2013; 14:1233-42.

25.Ornstein MC, Wood L, Elson P, et al. Clinical effect of dose escalation after disease progression in patients with metastatic renal cell carcinoma. Clin Genitourin Cancer 2017; 15:e275-80.

26.Ornstein MC, Pal SK, Wood LS, et al. Prospective phase II multi-center study of individualized axitinib (Axi) titration for metastatic renal cell carcinoma (mRCC) after treatment with PD-1 / PD-L1 inhibitors (abstract 4517). J Clin Oncol 2018; 36(suppl):4517.

Supplemental Data

Supplemental Table 1 Dose Modifications (Safety Population)

Pazopanib (n[ 554) Sunitinib (n[ 548) Dose Reduction(s), n (%) None 308 (56) 270 (49) Any 246 (44) 277 (51) 1 147 (27) 161 (29) 2 99 (18) 116 (21) Dose Interruption(s)‡7 Days, n (%) None 311 (56) 280 (51) Any 243 (44) 267 (48) 1 138 (25) 135 (25) 2 105 (19) 132 (24)

Supplemental Table 2 Logistic Regression Analysis for Baseline Factors Associated With a Response

OR (95% CI) P OR (95% CI) P CR/PR‡10 Months vs.<10 Months Pazopanib (n[ 76) Sunitinib (n[ 74) Age,_{<65 years vs. 65} years 0.789 (0.384-1.603) .591 1.163 (0.538-2.538) .814 Baseline KPS, 70-80 vs. 90-100 0.704 (0.291-1.697) .507 1.782 (0.693-4.714) .266 Number of metastatic sites,2 vs. 3 1.104 (0.543-2.235) .895 0.853 (0.384-1.895) .808

Number of organs involved, 1 vs._2 0.485 (0.162-1.378) .200 0.714 (0.174-2.642) .789 MSKCC risk category, favorable vs. intermediate vs. poor 0.594 (0.300-1.168) .141 0.673 (0.305-1.463) .366 CR/PR‡18 Months vs. <18 Months Pazopanib (n[ 34) Sunitinib (n[ 38) Age,<65 years vs. 65 years 0.788 (0.298-1.931) .738 1.040 (0.430-2.437) 1.000 Baseline KPS, 70-80 vs. 90-100 0.942 (0.330-3.099) 1.000 0.670 (0.249-1.904) .522 Number of metastatic sites,2 vs. 3 1.320 (0.541-3.106) .620 0.536 (0.190-1.368) .229

Number of organs involved, 1 vs. 2 1.473 (0.389-8.338) .801 0.596 (0.158-2.487) .568 MSKCC risk category, favorable vs. intermediate/ poor 0.784 (0.338-1.851) .668 0.426 (0.180-0.994) .048

Abbreviations: KPS¼ Karnofsky Performance Status; MSKCC ¼ Memorial Sloan Kettering Cancer Center.

COMPARZ Post Hoc Analysis of Pazopanib Responders

Supplemental Table 3 Efficacy by Baseline Metastatic Site

Baseline Metastatic Site

Pazopanib Sunitinib

Unadjusted HR (95% CI) Median PFS, Months (95% CI)

Bone Yes 8.1 (5.3-8.5) n¼ 110 8.3 (4.6-16.6) n¼ 85 1.221 (0.827-1.804) No 9.5 (8.3-11.1) n_{¼ 446} 10.9 (8.3-11.2) n_{¼ 468} 1.017 (0.860-1.203) Lung Yes 8.3 (8.3-8.5) n¼ 424 9.0 (8.2-11.0) n¼ 425 1.076 (0.904-1.279) No 11.1 (8.7-13.8) n¼ 132 11.2 (8.3-14.5) n¼ 128 1.023 (0.741-1.413) Kidney Yes 8.4 (6.8-11.1) n¼ 166 10.2 (8.2-13.6) n¼ 153 1.097 (0.817-1.473) No 8.4 (8.3-11.0) n¼ 390 9.1 (8.3-11.2) n¼ 400 1.051 (0.879-1.257)

Median OS, Months (95% CI)

Unadjusted HR (95% CI) Pa Bone Yes 17.4 (12.8-23.6) n_{¼ 110} 17.0 (10.9-21.1) n_{¼ 85} 0.935 (0.649-1.348) P ¼ .718 No 35.5 (28.2-NR) n¼ 446 31.5 (28.0-35.5) n¼ 468 0.901 (0.738-1.102) P ¼ .311 Lung Yes 27.2 (23.6-32.4) n¼ 424 27.5 (23.8-32.5) n¼ 425 0.988 (0.812-1.203) P ¼ .907 No 35.6 (31.2-NR) n¼ 132 31.5 (21.8-NR) n¼ 128 0.807 (0.550-1.183) P ¼ .271 Kidney Yes 27.6 (20.9-35.6) n¼ 166 29.8 (21.0-NR) n¼ 153 1.045 (0.757-1.444) P ¼ .789 No 31.6 (26.2-36.8) n¼ 390 28.2 (24.4-33.1) n¼ 400 0.906 (0.735-1.116) P ¼ .352 ORR, n (%) Difference, % (95% CI) Pb Bone Yes 23 (21) n¼ 110 13 (15) n¼ 85 6 (_{e5.2 to 16.4)} P ¼ .356 No 163 (37) n_{¼ 446} 147 (31) n_{¼ 468} 5 (e1.0 to 11.3) P ¼ .108 Lung Yes 153 (36) n¼ 424 117 (28) n¼ 425 9 (2.3-14.8) P ¼ .008 No 33 (25) n¼ 132 43 (34) n¼ 128 e9 (e19.6 to 2.4)P ¼ .136 Kidney Yes 44 (27) n¼ 166 33 (22) n¼ 153 5 (e4.4 to 14.3) P ¼ .359 No 142 (36) n¼ 390 127 (32) n¼ 400 5 (e1.9 to 11.3) P ¼ .177

Supplemental Table 4 Most Common AEs (‡5% Incidence in Either Treatment Group) Leading to Dose Modification(s)a Pazopanib (n[ 554) Sunitinib (n[ 548) Hypertension 71 (13) 37 (7) Fatigue 64 (12) 82 (15) Diarrhea 63 (11) 55 (10) Increased ALT 46 (8) 11 (2) Nausea 37 (7) 29 (5) PPE 35 (6) 68 (12) Vomiting 31 (6) 33 (6) Increased AST 30 (5) 7 (1) Neutropenia 13 (2) 44 (8) Thrombocytopenia 12 (2) 66 (12)

Data are presented as n (%).

Abbreviations: AE_{¼ adverse event; ALT ¼ alanine transaminase; AST ¼ aspartate} trans-aminase; PPE_{¼ palmar-plantar erythrodysesthesia.}

a_{Dose modi}

fications include dose reductions and dose interruptions resulting from AEs.

Supplemental Figure 1 Median Average Daily Dose by Dose Reduction Group for Pazopanib and Sunitinib. The Number of Patients on Study Treatment at<1, ‡1, ‡2, and ‡3 Years by Dose Reduction Group is Shown Below Each Panel

COMPARZ Post Hoc Analysis of Pazopanib Responders

Supplemental Figure 2 Incidence of Selected AEs by Dose Modification Group 2 ≥ e n o N Any 1 None ≥2 l l a r e v O Any 1

Dose reductions Dose interruptions ≥7 days 2 ≥ e n o N Any 1 None ≥2 l l a r e v O Any 1

Dose reductions Dose interruptions ≥7 days

Pazopanib Sunitinib Pazopanib Sunitinib

Pazopanib Sunitinib Pazopanib Sunitinib Pazopanib Sunitinib Pazopanib Sunitinib Pazopanib Sunitinib 2 ≥ e n o N Any 1 None ≥2 l l a r e v O Any 1

Dose reductions Dose interruptions ≥7 days

2 ≥ e n o N Any 1 None ≥2 l l a r e v O Any 1

Dose reductions Dose interruptions ≥7 days

2 ≥ e n o N Any 1 None ≥2 l l a r e v O Any 1

Dose reductions Dose interruptions ≥7 days 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 2 ≥ e n o N Any 1 None ≥2 l l a r e v O Any 1

Dose reductions Dose interruptions ≥7 days

2 ≥ e n o N Any 1 None ≥2 l l a r e v O Any 1

Dose reductions Dose interruptions ≥7 days 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 Patients, % Patients, % Patients, % Patients, % Patients, % Patients, % Patients, % e u g i t a F a e h r r a i D n o i s n e t r e p y H c i g o l o t a m e H T S A d e s a e r c n I T L A d e s a e r c n I PPE